Research Seminars at the Finnish Centre for Pensions Pension Information and Retirement Behaviour 22 March 2018 The seminar presents recent and on-going research from Finland and Norway on people’s pension knowledge. The seminar shows how information campaigns influence knowledge, attitudes and retirement plans. It also outlines how the potential impact of information gains persist over time.

1. Pension knowledge, impact
of economic incentives and
retirement intentions
Satu Nivalainen & Sanna Tenhunen
22.3.2018

2. Background
• Incentives matter
– economic incentives affect retirement timing (e.g. Brinch et al. 2014)
– people do react to the incentives in pension system, but only if they
know about them (e.g. Chan & Stevens 2008)
– the effect exists, but magnitude may remain low (e.g. Uusitalo &
Nivalainen 2013)
• Do people know about incentives in pension system?
– Gaps in knowledge of pension issues (e.g. Boeri & Tabellini, 2012)
– Key characteristics of pension system well known in Finland (Tenhunen
& Kuivalainen, forthcoming)
• Pension reform 2017 changed e.g. retirement age and effect
of advancing or postponing retirement on the amount of
pension
– Are people aware of new rules?
– Do incentives still matter?
222.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

3. Aim of the study
• Twofold:
1. To study
2. To study
322.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
Pension effectivness of
knowledge economic incentives
Economic retirement
incentives intentions

4. Data
• Collected by postal survey in 2016
• Questionnaire included info boxes on the reform
• Respondents
– N: 2 179 (response rate 56%)
– Finns
– aged 54-62 years
– non-retired
422.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

5. Pension knowledge
• Measured by three factors
– Has given an estimate of the amount of future pension
– Knows the effect of life-expectancy coefficient
– Has good knowledge on how continuing at work affects the amount
of accrued pension
522.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
47 54
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
No
41 59
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
64 36
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Yes
Yes
Yes
No
No

6. Pension knowledge
• In general, better knowledge among respondents of
– Older age (60-62)
– Higher education (tertiary)
– Higher pensionable income
– Entrepreneurs and upper white-collar workers
• Three measures of knowledge,
– only weakly correlated
» a fifth of respondents knew all three points, another fifth knew
none of them
– Using all three gives a fuller picture of knowledge
622.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

7. The effect of economic incentives
• Measured by three factors
– A person intends to postpone retirement due to life-expectancy
coefficient and delayed retirement benefit
– Delayed retirement benefit encourages to postpone retirement
– Early retirement deduction encourages to postpone retirement
722.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
16 84
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Yes No
57 43
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Yes No
46 54
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
NoYes

8. The impact of pension knowledge on the
effeciveness of economic incentives
822.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
• To find out if there is an impact, we modelled each three
claims on the effectiveness of economic incentives
– Controlling for
» demographics
» working sector (public/others)
» views on health and economic situation at retirement
» and the three measures of knowledge
– Logit model, presenting marginal effects

9. The impact of pension knowledge on the
effeciveness of economic incentives
Intends to
postpone
retirement due to
life-expectancy
coefficient and
delayed
retirement benefit
Delayed
retirement benefit
encourages to
postpone
retirement
Early retirement
deduction
encourages to
postpone
retirement
Has given an
estimate of the
amount of accrued
pension
0,026 0,043 0,011
Knows the effect of
life-expectancy
coefficient
0,015 -0,008 0,044
Has good knowledge
of how continuing at
work affects the
amount of accrued
pension
0,022 0,129*** 0,129***
922.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

10. The impact of pension knowledge on the
effeciveness of economic incentives
Intends to
postpone
retirement due to
life-expectancy
coefficient and
delayed
retirement benefit
Delayed
retirement benefit
encourages to
postpone
retirement
Early retirement
deduction
encourages to
postpone
retirement
Has given an
estimate of the
amount of accrued
pension
0,026 0,043 0,011
Knows the effect of
life-expectancy
coefficient
0,015 -0,008 0,044
Has good knowledge
of how continuing at
work affects the
amount of accrued
pension
0,022 0,129*** 0,129***
1022.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
Secondary or lowest
level tertiary education
Upper white-collar
workers
not concerned of their
work ability
not concerned of their
work ability
no spouse
assessing retirement
income to be moderate
or worse
Age 54-56 or 60-62
Secondary education
or higher
Entrepreneurs and
upper white-collar
workers
not concerned of their
work ability

11. Retirement intentions
• Based on the difference between
– self-assessed age of retirement and
– the lower limit of each person’s old-age retirement age
• Lower limit of old-age retirement age is defined by
– retirement age of each birth cohort
– or by person’s public sector personal retirement age (or other
supplementary pension with a different retirement age)
1122.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
Year of birth 1954 1955 1956 1957 1958 1959 1960 1961 1962
Retirement
age
63 y 63 y
3m
63 y
6 m
63 y
9 m
64 y 64 y
3 m
64 y
6 m
64 y
9 m
65 y

12. Retirement intentions
• Retirement intentions w.r.t. person’s own retirement age
1222.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
Exactly at
retirement
age
56%
After
27%
Before
retirement
age
17%

13. The effect of economic incentives on
retirement intentions
1322.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
• To find out if there is an impact, we modelled intentions
to retire before/exactly at/after own retirement age
– Controlling for
» demographics
» working sector (public/others)
» views on health and economic situation at retirement
» and the three measures of incentives
– Multinomial logit model, presenting marginal effects

14. The effect of economic incentives on
retirement intentions
Intends to retire
Before his own
old-age
retirement age
Exactly at his own
old-age
retirement age
After his own old-
age retirement
age
Intends to postpone
retirement due to
life-expectancy
coefficient and
delayed retirement
benefit
-0,062** -0,183*** 0,245***
Delayed retirement
benefit encourages
to postpone
retirement
-0,016 -0,049 0,065*
Early retirement
deduction
encourages to
postpone retirement
-0,024 0,046 -0,022
1422.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

15. The effect of economic incentives on
retirement intentions
Intends to retire
Before his own
old-age
retirement age
Exactly at his own
old-age
retirement age
After his own old-
age retirement
age
Intends to postpone
retirement due to
life-expectancy
coefficient and
delayed retirement
benefit
-0,062** -0,183*** 0,245***
Delayed retirement
benefit encourages
to postpone
retirement
-0,016 -0,049 0,065*
Early retirement
deduction
encourages to
postpone retirement
-0,024 0,046 -0,022
1522.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS
Age 60-62
Entrepreneurs, upper
and lower white-collar
workers
public sector workers
those not concerned of
their work ability
no spouse
assessing retirement
income to be rather
poor or poor
public sector workers
those concerned of
their work ability
assessing retirement
income to be moderate
or better

16. Conclusions
• Pension knowledge
– Considerable variation, others know, while some don’t
– Relation to effectiveness of incentives
» Knowing the effect of postponement on pension level
emphasizes the view that later retirement is motivated by
▪ delayed retirement benefit
▪ early retirement deduction
» Two other measures of pension knowledge did not affect
1622.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

17. Conclusions
• Retirement intentions
– There are both advancers and postponers, although majority plans
retirement at the earliest eligibility age
• The effect of economic incentives on intented retirement
age
– Intention to postpone increases
» when the person intends to postpone retirement due to life-
expectancy coefficient and delayed retirement benefit
» When delayed retirement benefit are seen as an effective
incentive
– Views on the early retirement benefit did not matter
1722.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS

18. Thank you for your attention
Research available at:
https://www.etk.fi/julkaisu/elaketietous-taloudellisten-kannustimien-
vaikuttavuus-ja-elakeaikeet/
For further information:
Satu.Nivalainen@etk.fi
Sanna.Tenhunen@etk.fi
1822.3.2018Sanna Tenhunen FINNISH CENTRE FOR PENSIONS